App safety

1. Secure the Code: Building a Secure Application

Mobile malware often taps vulnerabilities or bugs in the design and coding of the mobile applications they target. Recent research from Kindsight reported by Infosecurity shows that malicious code is infecting more than 11.6 million mobile devices at any given time, and the number of mobile malware samples is growing at a rapid clip, increasing by twentyfold in 2013.

Even before a vulnerability is exploited, attackers can obtain a public copy of an application and reverse engineer it. Popular applications are repackaged into “rogue apps” containing malicious code and are posted on third-party app stores to lure and trick unsuspecting users to install them and compromise their devices.

Enterprises should look for tools to aid their developers to detect and close security vulnerabilities and then harden their applications against reverse engineering and tampering. However, “consumer apps” still represent a threat as they may not undergo the appropriate hardening process; and if rogue applications, malware and enterprise apps share the same device, the threat is tangible.

2. Secure the Device: Detecting Compromised and Vulnerable Run-Time Environment

As secure as an application is, its security relies on the underlying device’s security. Jailbroken or rooted devices or the presence of rogue applications can represent an execution risk that may be allowed for certain enterprise apps but not for others.

Enterprises should look into ways to dynamically gauge the security of the underlying device. First, the mobile app sandbox, which is prevalent in modern mobile operating system design, must be intact. Rooting or jailbreaking the device breaks the underlying security model, and it is highly recommended to restrict these devices from accessing enterprise data. Furthermore, jailbreak technology is evolving rapidly to evade detection; coping with these mechanisms is essential to keeping up with these threats.

Mobile malware doesn’t always rely on the device being jailbroken, however. Excessive use of permissions to the mobile applications — which are granted by the user, often by default — can provide malware and rogue applications access to basic services (e.g., SMS) used to facilitate fraudulent activities.

Enterprises should consider up-to-date intelligence sources and application reputation services to track the tidal wave of applications — and their associated risk — as they enter mobile app stores on a daily basis. Using this data, application capabilities could be enabled or disabled based on the device risk profile.

3. Secure the Data: Preventing Data Theft and Leakage

When mobile applications access enterprise data, documents and unstructured information are often stored on the device. If the device is lost or when data is shared with nonenterprise applications, the potential for data loss is heightened.

Many enterprises are already looking into “remote wipe” capabilities to address stolen or lost devices. Mobile data encryption can be used to secure data within the application sandbox against malware and other forms of criminal access. To control application data sharing on the device, individual data elements should be encrypted and controlled.

4. Secure the Transaction: Controlling the Execution of High-Risk Mobile Transactions

Because mobile applications enable users to transact with enterprise services on the go, the risk tolerance for transactions will vary. For example, reading HR-related content may be deemed low risk versus the approval of a large payment to a new supplier.

Organizations should adapt an approach of risk-aware transaction execution that restricts client-side functionality based on policies that consider mobile risk factors such as device security attributes, user location, and the security of the network connection, among others.

Even when client-side transactions are allowed, enterprise applications can leverage an enterprise mobile risk engine to correlate risk factors such as IP velocity — access to the same account from two locations that are far apart over a short period — user access patterns and data access profiles. This approach extends the enterprise’s ability to detect and respond to complex attacks that can span multiple interaction channels and seemingly unrelated security events.